论文部分内容阅读
突发事件新闻报道数量的变化反映了突发事件自身发展态势以及媒体和公众对事件的反应的变化,是应急管理决策信息的重要来源。本文基于隐马尔可夫模型对突发事件新闻报道的爆发性进行了建模,以反映突发事件新闻报道数量的变化趋势。本文还提出了使用时间序列聚类算法去识别突发事件新闻报道数量的演化模式。对28起突发事件新闻报道的实验分析表明,本文提出的爆发性建模方法能够详细和准确地描述突发事件新闻报道数量的变化。此外,本文从这28起突发事件新闻报道数量的时间序列中识别了四类演化模式并分析了每一类的特征。
Changes in the number of news reports of emergencies reflect the development of emergencies themselves and the changes in media and public reactions to incidents, which are important sources of emergency management decision-making information. Based on the hidden Markov model, this article models the outbreak of emergency news reports to reflect the changing trend of the number of emergency news stories. This paper also proposed the use of time series clustering algorithm to identify the evolution of the number of emergency news reports. The experimental analysis of 28 incidents news reports shows that the explosive modeling method proposed in this paper can describe the changes of incidents news coverage in detail and accurately. In addition, the paper identifies four types of evolutionary patterns from the time series of 28 incident news reports and analyzes the characteristics of each category.